Mailing List Archive: S50n problems w/ stacking,packet drops

We've been demoing some S50n units and have been having some problems which I
have a hard time seeing are "normal", but wanted to get some outside
perspective on things.

Our first pair of demo units wouldn't stack properly. They pretty much
wouldn't see each other.

These were replaced and the next set seemed to stack fine. Then one of the
units started having a problem where it would loose its config and need reset
so it would boot properly if it was left unplugged for more than an hour or so.

After we got over these issues we tried to put them in production and ended up
having really bad packet loss to hosts even directly connected to the switches
(~60% in some cases). We weren't pushing a lot of bandwidth; our usual
traffic is probably < 6gbps across an entire 48-port switch. The switch was
doing primarily only layer 2 switching; the only layer 3 stuff it would have
touched would have been management request to itself.

Has anyone else had experiences like we've had? I'm assuming this is
abnormal, but just wanted some outside experiences.

-Doug

I would say that it's definitely not normal and worth calling your rep about. It's probably worth mentioning that I've never considered their, or anyone else's for that matter, stacking technology reliable enough to use in production. In environments that require more ports than an individual S50n I've usually moved up to bigger switches like the C or E series or if that is too big a step connected the S50s and used 1g or 10g as uplink ports to bridge vlans depending on bandwidth requirements or on occasion a 2410 spanning out to a number of S50s.

I'm sorry if this is not the answer you were looking for ... just my 2 bits.

David

On Feb 26, 2010, at 10:26 AM, Doug Warner wrote:

> We've been demoing some S50n units and have been having some problems which I
> have a hard time seeing are "normal", but wanted to get some outside
> perspective on things.
>
> Our first pair of demo units wouldn't stack properly. They pretty much
> wouldn't see each other.
>
> These were replaced and the next set seemed to stack fine. Then one of the
> units started having a problem where it would loose its config and need reset
> so it would boot properly if it was left unplugged for more than an hour or so.
>
> After we got over these issues we tried to put them in production and ended up
> having really bad packet loss to hosts even directly connected to the switches
> (~60% in some cases). We weren't pushing a lot of bandwidth; our usual
> traffic is probably < 6gbps across an entire 48-port switch. The switch was
> doing primarily only layer 2 switching; the only layer 3 stuff it would have
> touched would have been management request to itself.
>
> Has anyone else had experiences like we've had? I'm assuming this is
> abnormal, but just wanted some outside experiences.
>
> -Doug
>
> _______________________________________________
> force10-nsp mailing list
> force10-nsp@puck.nether.net
> https://puck.nether.net/mailman/listinfo/force10-nsp

_______________________________________________
force10-nsp mailing list
force10-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/force10-nsp

Mailing List Archive

Mailing List Archive

Attached Files: